LiveKit: offline framework-adapter simulation

Twin: examples/sdk_framework_adapter_livekit_run_session_promotion.py · emits agent-learning.run.v1 · offline, no credentials. A coding agent can complete this page from the frontmatter alone.

1. What you are testing

LiveKit coverage in the kit is probe-promoted, and this page is deliberately offline: it tests the agent-session adapter, not a live room. The twin, examples/sdk_framework_adapter_livekit_run_session_promotion.py, builds a local LocalLiveKitAgentSession with two surfaces: a respond(text) path that returns content with no room-session evidence, and an async run_session(payload) path that returns verified evidence only when the adapter passes metadata.framework == "livekit" in the payload dict. Promotion selects the session entrypoint with evidence; the text path is recorded as weak.

The failure class this catches is session-shape mismatch: a voice agent harness that drives the text surface can look healthy while never exercising the session contract a real room would use. Pinning the adapter shape offline means the behavioral contract is already proven before any live infrastructure enters the picture.

The run manifest, examples/framework_livekit_manifest.json, drives the same adapter from the CLI. It targets the factory framework_shims.py:build_livekit_agent with trace_runtime: true and replays a framework_trace environment whose span is agent.respond. Everything runs on the local_text engine in one turn: offline, deterministic, no provider keys.

2. Run it

CLI (the required_env key is CI metadata for this offline manifest — any placeholder value satisfies it):

AGENT_LEARNING_MULTI_FRAMEWORK_EXAMPLE_KEY=local-example \
agent-learn run examples/framework_livekit_manifest.json \
  --output artifacts/framework-livekit.json

SDK, same operation (export the same placeholder env first):

import asyncio
from fi.alk import simulate

result = asyncio.run(
    simulate.run_manifest_file("examples/framework_livekit_manifest.json")
)
assert result["kind"] == "agent-learning.run.v1"

3. What you built

Postcondition (machine-checkable — the same check the docs gate pattern uses):

python -c "import json; p=json.load(open('artifacts/framework-livekit.json')); assert p['kind']=='agent-learning.run.v1', p['kind']; print('ok')"

The artifact carries status, a per-turn transcript for the framework-owner scenario, the evaluation report, and the framework runtime trace evidence the adapter extracted — including the agent.respond span the manifest replays. It is a replayable record, not a log line: the same file feeds baseline, compare, and replay.

4. When it fails

Symptom	First-mile class	Doctor check
vendored import failed	infra	agent-learn doctor → summary.missing_engine_modules
missing required environment variable(s)	config fault	set the placeholder key shown above; agent-learn doctor → summary.public_boundary_passed confirms the install surface
transcript shows the weak respond(text) path (no session evidence)	behavior regression	re-run the twin promotion and compare run_session evidence against the text fallback

5. Prove it / keep it

The twin is admitted by the framework_adapter_probe_readiness release gate, so every agent-learn release-check re-executes this exact session promotion path — the page stays true or the release fails. To keep your own LiveKit agent honest, promote the run artifact into a regression baseline with the baseline / promote-to-regression / compare command family, then wire the manifest into CI. Live LiveKit sessions (real rooms, real audio) are an opt-in lane (see ROADMAP.md: voice lane rungs) — this page stays on the offline golden path. The reader's job here is maintenance of a living proof, not a one-off demo.